Parkinson's disease (PD) is characterized by both motor and non-motor symptoms (cognitive impairment, affective disorder, and other clinical features). Data from experimental animal models and patients with PD indicate that the manifestations of this disease cannot be attributed to isolated dysfunction of the basal ganglia. Rather, the highly localized loss of nigral dopamine cells is associated with a broad, spatially distributed set of functional abnormalities involving cortico-striato-pallido-thalamocortical (CSPTC) loops and related pathways. By quantifying the activity of spatially distributed (large-scale) functional brain networks, comprising multiple interconnected brain regions, modern techniques of image-based analysis can provide valuable information concerning the widespread circuit abnormalities that underlie neurodegenerative disorders such as PD. The investigators at the Center for Neurosciences at The Feinstein Institute, led by Dr. Eidelberg, have pioneered the use of functional brain imaging and network analysis for the study of PD and related neurodegenerative diseases. Because of the noise inherent in small signals analyses of this sort, we have emphasized rigorous validation of the disease-related functional patterns from both statistical and empiric standpoints. Indeed, high levels of measurement precision are needed before quantitative network measures can be considered as potential biomarkers of the disease process and its response to treatment. In this proposal, we seek to take this approach to a new level by employing rigorously validated PD-related networks to address a number of vital issues that impact heavily on the care of today's PD patients. Project 1 addresses the serious clinical problem of levodopa-induced dyskinesias, which ultimately affect nearly all PD patients. Project 2 examines the network basis for individual differences in the cognitive response to dopaminergic treatment with a view to predicting which patients will develop untoward cognitive side effects under different treatment conditions. Project 3 aims to establish the feasibility of a new network-based algorithm for providing earlier and more accurate differential diagnosis than is currently possible.